Personalized search based on user intention through the hierarchical phrase vector model
Gun-Woo Park, JinGi Chae, Dae Hee Lee, Sang-Hoon Lee · 2008
There are many kinds of personalizing approaches in the area of web information retrieval. But it is still unclear whether personalization is consistently effective on different queries for different users, and under different search contexts. In this paper, we study this problem and propose a personalized search approach that can easily extend a conventional search engine. We present an intelligent relevance-evaluation framework for user Intention-based personalized search based on web-mining and machine learning approaches. Users can navigate through the prior user's intention by their own needs. This is especially useful for polysemous and poor queries. By analyzing the results, we reveal that there is an unique representation of intention under different queries, contexts and users. Furthermore, we reveal that this knowledge is very important in improving retrieval performance by filtering the results, recommending a new query, and distinguishing user's characteristics. Proposed personalized search engine HPS(Hierachical Phrase Search) can provide more predictive information by clearly identifying the searcher's intention at query-time.